Table 2. Multivariate regression between predictors and slope of the trends in the perceived prevalence of smoking in movies1.
| Predictors | Age cohort (years) | |||||
|---|---|---|---|---|---|---|
| 11 | 12 | 13 | 14 | 15 | 16 | |
| Demographics | ||||||
| Male (vs female) | −0.03 | −0.01 | ||||
| Nonwhite (vs white) | −0.00 | 0.02 | −0.03 | |||
| Parent education | 0.00 | 0.00 | 0.01* | 0.00 | 0.02 | 0.00 |
| Level of urbanization | 0.00 | |||||
| Attitudes toward tobacco companies | ||||||
| Get too much blame (reversed) | −0.00 | 0.00 | ||||
| Change in agreeing “get too much blame (reversed)” | 0.03 | |||||
| Make too much money from teens | −0.01 | −0.01 | 0.00 | −0.01 | −0.04 | |
| Change in agreeing “make too much money from teens” | 0.04* | |||||
| Get teens smoking | −0.01 | −0.01 | −0.01 | 0.02 | ||
| Social environment | ||||||
| Number of smoking close friends | −0.02 | −0.01 | 0.01 | −0.01 | −0.03* | −0.00 |
| Living with smoking siblings | −0.05 | −0.06* | −0.04 | −0.03 | ||
| Living with smoking parents | −0.01 | 0.01 | −0.03 | 0.01 | −0.06 | |
| Change in living with smoking parents | −0.21* | |||||
| Home smoking restriction score | −0.00 | 0.00 | 0.00 | 0.01 | 0.02 | |
1
Predictors not significantly associated with either intercepts or slopes in bivariate analysis (p≥ 0.10) were not included in the models. Demographic predictors were adjusted for each other; other variables were adjusted for all variables with significant bivariate associations with either the intercept or the slope.
*
p<0.01.